Man is not a big rat: concerns with traditional human risk assessment of phthalates based on their anti-androgenic effects observed in the rat foetus.

Phthalates provide one of the most documented example evidencing how much we must be cautious when using the traditional paradigm based on extrapolation of experimental data from rodent studies for human health risk assessment of endocrine disruptors (EDs). Since foetal testis is known as one of the most sensitive targets of EDs, phthalate risk assessment is routinely based on the capacity of such compounds to decrease testosterone production by the testis or to impair masculinization in the rat during foetal life. In this paper, the well-established inhibiting effects of phthalates of the foetal Leydig cells function in the rat are briefly reviewed. Then, data obtained in humans and other species are carefully analysed. Already in January 2009, using the organotypic culture system named Fetal Testis Assay (FeTA) that we developed, we reported that phthalates might not affect testosterone production in human foetal testes. Several recent experimental studies using xenografts confirm the absence of detectable anti-androgenic effect of phthalates in the human foetal testes. Epidemiological studies led to contradictory results. Altogether, these findings suggest that phthalates effects on foetal Leydig cells are largely species-specific. Consequently, the phthalate threshold doses that disturb foetal steroidogenesis in rat testes and that are presently used to define the acceptable daily intake levels for human health protection must be questioned. This does not mean that phthalates are safe because these compounds have many deleterious effects upon germ cell development that may be common to the different studied species including human. More generally, the identification of common molecular, cellular or/and phenotypic targets in rat and human testes should precede the choice of the toxicological endpoint in rat to accurately assess the safety threshold of any ED in humans.